# tests/factor/test_fundamental_batch.py
"""财务因子批端到端: 合成行情库 + 合成静态域 → batch_eval 出 IC 指标."""
import sqlite3
import sys, os
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..")))
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..", "vnpy_v4.4.0")))

import pytest

from sanguo_factor import alpha_datasets  # 挂载量价因子
from sanguo_factor import fundamental_library  # noqa: F401  挂载财务因子(import 即注册)
from sanguo_factor.batch_eval import run_batch_eval
from sanguo_factor import eval_store

@pytest.fixture(autouse=True)
def _ensure_fundamental_registered():
    """其它测试模块清空 _REGISTRY 后只重挂 alpha/builtin(顺序依赖前科),
    这里逐测试幂等重注册财务因子,保证本模块与顺序无关."""
    fundamental_library._register_all()


_DDL = """
CREATE TABLE dbbardata(
  symbol TEXT, exchange TEXT, datetime TEXT, interval TEXT,
  volume REAL, turnover REAL, open_interest REAL,
  open_price REAL, high_price REAL, low_price REAL, close_price REAL)
"""

_STOCKS = [("600000", "SSE"), ("000001", "SZSE"), ("300001", "SZSE")]


@pytest.fixture(scope="module")
def db(tmp_path_factory):
    """3 只 × 2022-06~2024-01 合成日线(与合成财务域同代码;趋势构造截面差异)."""
    alpha_datasets.mount_all()
    p = tmp_path_factory.mktemp("fb")
    db = str(p / "qt.db")
    conn = sqlite3.connect(db)
    conn.execute(_DDL)
    import pandas as pd
    days = pd.bdate_range("2022-06-01", "2024-01-05")
    drift = {"600000": 0.0012, "000001": -0.0006, "300001": 0.0004}
    base = {"600000": 10.0, "000001": 5.0, "300001": 20.0}
    for i, day in enumerate(days):
        d = day.strftime("%Y-%m-%d")
        rows = []
        for sym, ex in _STOCKS:
            px = base[sym] * (1.0 + drift[sym]) ** i
            rows.append((sym, ex, f"{d} 00:00:00", "d", 100.0, px * 100.0, 0,
                         px, px, px, px))
        conn.executemany("INSERT INTO dbbardata VALUES(?,?,?,?,?,?,?,?,?,?,?)", rows)
    conn.commit()
    conn.close()
    return db


def test_fundamental_end_to_end(db, synthetic_static, tmp_path):
    eval_db = str(tmp_path / "fund_eval.db")
    out = run_batch_eval(
        factor_names=["fund_roe_ttm", "fund_tacc", "fund_ep_ttm",
                      "fund_forecast_type"],
        start="2023-02-01", end="2023-12-31",
        eval_db=eval_db, label="fund_t", cfg=None, vnpy_db_override=db,
        fund_data_dir=synthetic_static,
    )
    assert out["factors_done"] == 4
    assert out["errors"] == []
    for name in ("fund_roe_ttm", "fund_tacc", "fund_ep_ttm", "fund_forecast_type"):
        m = eval_store.get_detail(eval_db, out["run_id"], name)["metrics"]
        assert "error" not in m, f"{name}: {m.get('error')}"
        # 截面 3 只 → 逐日 rank IC 可算(预告因子 2023-07-15 后才有效, count 少但非零)
        assert isinstance(m["1"]["ic_mean"], float) or m["1"]["ic_mean"] is None
        assert m["1"]["count"] >= 0


def test_mixed_batch_price_plus_fundamental(db, synthetic_static, tmp_path):
    """量价+财务同批混跑: 白名单 join 只在含财务因子时发生,互不干扰."""
    eval_db = str(tmp_path / "mix_eval.db")
    out = run_batch_eval(
        factor_names=["ma_20", "fund_sue_np"],
        start="2023-02-01", end="2023-06-30",
        eval_db=eval_db, label="mix_t", cfg=None, vnpy_db_override=db,
        fund_data_dir=synthetic_static,
    )
    assert out["errors"] == []
    rows = eval_store.get_rows(eval_db, out["run_id"])
    assert {r["factor"] for r in rows} == {"ma_20", "fund_sue_np"}
    # SUE 在 2023-06-30 前可见的报告期均无完整 8 期窗 → 全 NaN → eliminated 非 error
    m = eval_store.get_detail(eval_db, out["run_id"], "fund_sue_np")["metrics"]
    assert "error" not in m


def test_fundamental_p1_end_to_end(db, synthetic_static, tmp_path):
    """P1 批端到端: VSIG(16 季窗)+EV 群(分母含 ev_ex_mv)+兑现差走引用瘦身 join."""
    eval_db = str(tmp_path / "fund_p1_eval.db")
    out = run_batch_eval(
        factor_names=["fund_vsig", "fund_ebit_over_ev", "fund_forecast_beat",
                      "fund_profit_streak"],
        start="2023-02-01", end="2023-12-31",
        eval_db=eval_db, label="fund_p1", cfg=None, vnpy_db_override=db,
        fund_data_dir=synthetic_static,
    )
    assert out["factors_done"] == 4
    assert out["errors"] == []
    for name in ("fund_vsig", "fund_ebit_over_ev", "fund_forecast_beat",
                 "fund_profit_streak"):
        m = eval_store.get_detail(eval_db, out["run_id"], name)["metrics"]
        assert "error" not in m, f"{name}: {m.get('error')}"
    # EV 群/持续盈利全截面有值 → IC 样本点非零;vsig 仅 2/3 股非空(B 窗含
    # NaN)→ <3 截面 IC 记 0(n>=3 引擎下限),路径本身已被无错误评估锁死
    for name in ("fund_ebit_over_ev", "fund_profit_streak"):
        m = eval_store.get_detail(eval_db, out["run_id"], name)["metrics"]
        assert m["1"]["count"] > 0, name


def test_fundamental_p1b_end_to_end(db, synthetic_static, tmp_path):
    """P1-B 批端到端: vb 日频域 + dividend 事件流走引用瘦身域级加载."""
    eval_db = str(tmp_path / "fund_p1b_eval.db")
    out = run_batch_eval(
        factor_names=["fund_ep_vb", "fund_send_total_12m"],
        start="2023-08-01", end="2023-12-31",
        eval_db=eval_db, label="fund_p1b", cfg=None, vnpy_db_override=db,
        fund_data_dir=synthetic_static,
    )
    assert out["factors_done"] == 2
    assert out["errors"] == []
    for name in ("fund_ep_vb", "fund_send_total_12m"):
        m = eval_store.get_detail(eval_db, out["run_id"], name)["metrics"]
        assert "error" not in m, f"{name}: {m.get('error')}"
    # ep_vb 仅 2/3 股非空(C 亏损 pe=NaN)→ <3 截面 IC 记 0(n>=3 引擎下限,
    # 与 P1 vsig 同形态);域级加载/join 路径本身已被无错误评估锁死
    m = eval_store.get_detail(eval_db, out["run_id"], "fund_ep_vb")["metrics"]
    assert m["1"]["count"] >= 0


def test_pure_price_batch_untouched(db, tmp_path):
    """纯量价批不传静态域照常跑(回归: 未新增强依赖)."""
    eval_db = str(tmp_path / "px_eval.db")
    out = run_batch_eval(
        factor_names=["roc_5"], start="2023-02-01", end="2023-03-31",
        eval_db=eval_db, label="px_t", cfg=None, vnpy_db_override=db,
    )
    assert out["errors"] == []
    assert out["factors_done"] == 1